Re: Unicode...

[Paul A Vixie <vixie@xxxxxxxxxxxxxx>]

There are two different bit-oriented questions being discussed today, and
I think it would be beneficial to think of them separately.  On the one
hand, there is the question: "what code should the body use - BCD or BAUDOT?"
On the other hand, there is the question, "what bit patterns should be
passable through the low-level transport?"

These are separate questions because the first one has multiple (infinite?)
answers and we can add Header:'s to describe the code (and structure and
so on) of the body.  I suppose that argument is possible on whether we
ought to force everyone to use a single structure mechanism and a single
encoding scheme (probably Unicode but at least ISO 8897, or 10616161???),
or whether we should add more grist to "assigned numbers" by creating a set
of possible encoding coupons, or whether we should have the header describe
the complete encoding in NP, or, well, it's getting silly.

The second question, transport width, is sticky only because there are 
already some low-level (non-Internet) mail transports that can do 8 bits,
and for those and for future Internet mail transport protocols, noone here
wants to limit the user agents to the current 7 bits.  

It looks to me like the findings of this WG are:

	(1) we need an 8-bit transport protocol, and we have 1,400
		different ways of perverting (I mean "extending")
		SMTP for this, just ask us for the list;

	(2) we need more things like RFC934 so we can send complex
		objects as electronic mail (binary programs, multi-
		media, pornography, multiple-attachments, and so on).

These findings are not in conflict with each other.

Cheers,

Paul Vixie
DEC Western Research Lab	<vixie@wrl.dec.com>
Palo Alto, California		...!decwrl!vixie